In order to evaluate the ability of three types of extracts to explain the ecotoxicological risk of treated municipal sewage sludges, the OECD 208A germination test was applied using three plants (Lolium perenne L., Brassica rapa L., and Trifolium pratense L.). Three equivalent batches of sludge, remained as dewatered sludge, composted with plant remains and thermally dried, from an anaerobic waste water treatment plant were separated. Samples from these three batches were extracted in water, methanol, and dichloromethane. Plant bioassays were performed and the Germination Index (GI) for the three plants was evaluated once after a period of 10 days. Germination in extracts was always lower than the respective controls. The germination in composted sludge (GI 40.9-86.2) was higher than the dewatered (GI 2.9-45.8), or thermally dried sludges (GI 24.6-64.4). A comparison of the germination between types of extracts showed differences for dewatered sludge with the three plants, where the water and methanol extracts had significantly lower germination than the dichloromethane extract. A higher half maximal effective concentration (EC50) in composted extracts was established, mainly in the water extract (EC50 431-490 g kg(-1)). On the contrary, the germination was strongly inhibited in the water extract of the dewatered sludge (EC50 14 g kg(-1)). The germination was positively correlated with the degree of organic matter stability of the parent sludge, and an inverse correlation was detected for total nitrogen, hydrolysable nitrogen and ammonium content. It is concluded that the phytotoxic effect of the water extract is more closely related to hydrophilic substances rather than lipophilic ones, and care must be taken with dewatered sludge application, especially with their aqueous eluates. Results obtained in this work show the suitability of the use of sludge extracts in ecotoxic assays and emphasize the relevance of sewage sludge stabilization by post-treatment processes.